Mitochondrial-associated endoplasmic reticulummembranes (MAM) form innate immune synapsesand are targeted by hepatitis C virusStacy M. Horner, Helene Minyi Liu, Hae Soo Park, Jessica Briley, and Michael Gale, Jr.1

Department of Immunology, University of Washington School of Medicine, Seattle, WA 98195

Edited* by Francis V. Chisari, The Scripps Research Institute, La Jolla, CA, and approved July 19, 2011 (received for review June 22, 2011)

RIG-I is a cytosolic pathogen recognition receptor that engagesviral RNA in infected cells to trigger innate immune defensesthrough its adaptor protein MAVS. MAVS resides on mitochondriaand peroxisomes, but how its signaling is coordinated amongthese organelles has not been defined. Here we show that a majorsite of MAVS signaling is the mitochondrial-associated membrane(MAM), a distinct membrane compartment that links the endoplas-mic reticulum to mitochondria. During RNA virus infection, RIG-I isrecruited to the MAM to bind MAVS. Dynamic MAM tethering tomitochondria and peroxisomes then coordinates MAVS localiza-tion to form a signaling synapse between membranes. Impor-tantly, the hepatitis C virus NS3/4A protease, which cleavesMAVS to support persistent infection, targets this synapse forMAVS proteolysis from the MAM, but not from mitochondria, toablate RIG-I signaling of immune defenses. Thus, the MAM medi-ates an intracellular immune synapse that directs antiviralinnate immunity.

interferon | IPS-1

MAVS, formerly also called VISA/CARDIF/IPS-1, is es-sential for innate immunity to RNA viruses (1, 2). In ac-

cordance with the HUGO Gene Nomenclature Committee–approved nomenclature, we now refer to this molecule asMAVS. During acute infection by hepatitis C virus (HCV) andother RNA viruses, viral RNA is recognized in the host cell asnonself by the cytosolic pathogen recognition receptor RIG-I,resulting in a signaling cascade that induces innate antiviral im-munity by triggering the production of type I IFN and proin-flammatory cytokines from infected tissue (3, 4). The signals thatdrive this response are relayed through MAVS, which is localizedto both peroxisomes and mitochondria, for assembly of a mac-romolecular signaling complex (5, 6). The importance of MAVSin innate immunity is underscored by the fact that HCV, which issensed by RIG-I in hepatocytes (7), uses its multifunctional NS3/4A protease to cleave MAVS from membranes to evade RIG-Isignaling (8–11). This innate immune evasion strategy contrib-utes to HCV’s ability to establish chronic infection in nearly 200million people worldwide and provides a foundation for liverdisease. The NS3/4A protease has been reported by us andothers to cleave MAVS on the outer mitochondrial membrane(OMM); however, its actions in HCV replication occur on theendoplasmic reticulum (ER) (10, 12). Interestingly, the ERcontains a specialized subdomain, the mitochondrial-associatedmembrane (MAM), that physically connects the ER to mito-chondria and provides a mitochondrial–ER axis that compart-mentalizes both stress and metabolic signaling (13, 14). TheMAM plays a central role in calcium signaling and phospholipidtransport to mitochondria (15). In addition, the MAM has re-cently been implicated as a platform for inflammasome signaling,suggesting that it might play a role in infection and immunity(16). We hypothesized that the MAM may coordinate MAVSsignaling of innate immunity from peroxisomes and mitochon-dria, and thus that MAM-localized MAVS could represent theactual substrate target of the HCV NS3/4A protease to facilitateviral immune evasion.

ResultsWe assessed MAVS cellular distribution by confocal microscopyof human hepatoma (Huh7) cells expressing MAVS-mEGFP. Inaddition to the previously described mitochondrial and peroxi-somal localization of MAVS (5, 6), our higher-resolution analy-ses also revealed MAVS staining directly adjacent to and aroundmitochondria (Fig. 1A). The MAM is known to link the ER tomitochondria (14), and we assessed the possibility that it alsocould make contacts with peroxisomes, which are derived fromER membranes (17). Indeed, confocal microscopy of Huh7 cellsexpressing mEGFP-PSS-1, a MAM-enriched marker protein (18,19), revealed an interconnected MAM network, distinct from theER, between mitochondria and peroxisomes (Fig. S1 A and B).Thus, we biochemically assessed the cellular distribution ofMAVS and various intracellular organelle marker proteins byfractionating Huh7 cells to isolate the MAM from mitochondriaand microsomes, the latter of which contains light membranesof the ER and plasma membrane (Fig. S2A) (18). MAVS waspresent in Huh7 cells within bothMAMandmitochondria, as wellas in “F1”, a membrane fraction of lighter density than MAMalone that contains bothMAMand peroxisomes (Fig. 1B), but notin microsomal fractions (Fig. S2B). The MAM-localized MAVScofractionated with a previously described MAM-enrichedmarker, FACL4 (Fig. 1B and Fig. S2 B–D) (15, 18). The MAMfractions were largely devoid of both peroxisomal (Pex19) andmitochondrial markers (Cox-1 and VDAC1) (Fig. 1B and Fig.S2B), whereas standard mitochondrial fractions contained all ofthese markers (Fig. S2B). In addition, MFN2, a MAM- and mi-tochondrial-localized protein that tethers the ER to mitochondriaand interacts with MAVS (20, 21), cofractionated with MAVSand was present in both MAM and mitochondria (Fig. S2C).To verify biochemical identification of the MAM localization

of MAVS, we performed confocal microscopy of endogenousMAVS in Huh7 cells. This revealed MAVS staining directly ad-jacent to mitochondria and partially overlapping with three dif-ferent MAM-enriched proteins: FACL4, sigma-1 receptor (Sig-1R), and mEGFP-PSS-1 (Fig. 1C and Figs. S2D and S3) (15, 18).This MAVS distribution pattern at the MAM–mitochondrial axiswas apparent in cells that were either mock-infected or infectedwith Sendai virus (SenV), which activates RIG-I signaling andaccurately models HCV activation of the RIG-I pathway (Fig.S3) (7, 10). Thus, MAVS is localized to the MAM, peroxisomes,and mitochondria, likely residing at the junctions between theseorganelles.The HCV NS3/4A protease complex is localized within

infected cells to intracellular membranes and regions near mi-

Author contributions: S.M.H., H.M.L., and M.G. designed research; S.M.H., H.M.L., H.S.P.,and J.B. performed research; S.M.H., H.M.L., and M.G. analyzed data; and S.M.H. and M.G.wrote the paper.

The authors declare no conflict of interest.

*This Direct Submission article had a prearranged editor.1To whom correspondence should be addressed. E-mail: mgale@uw.edu.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1110133108/-/DCSupplemental.

14590–14595 | PNAS | August 30, 2011 | vol. 108 | no. 35 www.pnas.org/cgi/doi/10.1073/pnas.1110133108

Dow

nloa

ded

by g

uest

on

Apr

il 26

, 202

1

tochondria by membrane-anchor domains within NS3 and theNS4A cofactor (22), suggesting that MAM localization of NS3/4A could govern its dual distribution between the ER and mi-tochondria. Therefore, we assessed the distribution of NS3/4Atransiently expressed in HeLa cells that coexpress mEGFP-PSS-1. We found that NS3 and NS4A fractionated with MAM, F1(representing peroxisomes with MAM), and microsomes (rep-resenting the ER and HCV replication complex; described in ref.12), with a small amount of NS4A also detected in the mito-chondria (Fig. 2A). Confocal microscopy confirmed NS3/4A lo-calization adjacent to peroxisomes and mitochondria, both ofwhich are sites of MAVS localization (Fig. 2C) (5, 6). In addi-tion, confocal microscopy revealed that in cells propagating aself-replicating HCV RNA genome (a model of persistent HCVinfection; Huh7-HCV K2040 cells) (3), both NS3 and NS4Awere present at MAM–mitochondrial junctions (Fig. 2B). InHCV-infected Huh7.5 cells (3), both NS3 and NS4A were as-sociated with the MAM either (i) at junctions with mitochondria(Fig. 2 D and E, large arrows) or (ii) independent of mito-chondria (Fig. 2 D and E, small arrows), similar to what wasobserved in Huh7-HCV K2040 cells (compare Fig. 2 B and E).Thus, biochemical fractionation and direct imaging of cells es-tablish that NS3/NS4A localizes to the junctions among MAM,mitochondria, and peroxisomes during HCV infection and per-sistent viral RNA replication.In cells replicating HCV RNA, NS3 was found in both MAM

and microsomal fractions (Fig. 3 A and B and Fig. S2C). MAVSwas cleaved during HCV replication by NS3/4A and accumulatedwithin the cytosolic fraction (Fig. 3B, lane 8 and Fig. S2C), asdescribed previously (10). Importantly, we detected cleavedMAVS in the MAM, but not in the mitochondrial fraction, duringHCV RNA replication (Fig. 3B, compare lanes 4, 6, and 8; Fig.S2C). The residual cleaved MAVS present in the MAM likely

reflects the homotypic interaction of MAVS with uncleaved pro-tein and/or binding to other MAM-associated signaling cofactors,with the cleaved MAVS in the cytosol likely representing bothMAM and peroxisomal-MAVS cleavage by NS3/4A. Surprisingly,neither the mass nor the relative amount of mitochondrial MAVSwas affected by the presence of HCV proteins (Fig. 3B, lanes 5and 6, with markers of these samples shown in Fig. 3A; see alsoFig. S2C). We note that the lack of MAVS cleavage from mito-chondria by NS3/4A differs from findings in previous studies byus and others in which analysis of crude mitochondrial fractionsled to the conclusion that MAVS was cleaved from the OMM(10, 23). We have now found that such mitochondrial fractionsindeed contain MAM, explaining these differences (Fig. S2B).Moreover, confocal microscopy analysis of cells expressing NS3/4A identified residual mitochondrial MAVS staining, thus de-fining a pool of uncleaved mitochondrial-associated MAVS (Fig.S4). Therefore, MAVS is cleaved by NS3/4A on the MAM, butnot on mitochondria, during HCV infection.Because MAM, but not mitochondrial MAVS, is the pro-

teolytic target of NS3/4A, we evaluated the possibility that theMAM-localized MAVS is capable of transducing RIG-I signal-ing. We assessed MAM fractions from Huh7 and Huh7-HCVK2040 cells for the presence of RIG-I and found that a portionof RIG-I was redistributed from cytosol into the MAM, but notinto the mitochondria, in cells chronically replicating HCV RNA(Fig. 3 A and B and Fig. S2C). This redistribution was not due tocytosol contamination of these fractions or to changes in MAMlevels during HCV RNA replication, because the cytosol markertubulin was absent from the MAM and microsomes, and MAMmarker levels were unchanged by HCV RNA replication (Fig. 3A and C and Fig. S2C). RIG-I levels are known to increase afterRIG-I pathway activation (9) and were accordingly increasedduring the course of SenV infection (Fig. 3C). RIG-I levels also

A

1168237

823726

MA

M &

Pex

(F1)

MA

M

Mito

chon

dria

82

calnexin

Cox-1

FACL4

Pex19

MAVS

CB

TOM20PMP70 egreMSVAM

Zoom

Merge

Distance

%In

tens

ity%

Inte

nsity

1

2

1

2

MAVS

FACL4

Mito.

Merge

Zoom

%In

tens

ity

Distance

Merge

MAVSFACL4 Mito. FACL4

MAVS Mito.

Fig. 1. MAVS is localized to MAM, mitochondria, and peroxisomes. (A) Confocal microscopy analysis of immunostained Huh7 cells expressing MAVS-mEGFPfor PMP70 (peroxisomes) and TOM20 (OMM). (B) Immunoblot analysis of fractionated Huh7 cells. F1, containing MAM and peroxisomes, is labeled as MAM &Pex. Fractionation markers: calnexin, ER; Cox-1, mitochondria; FACL4, MAM; Pex19, peroxisomes. (C) Confocal microscopy analysis of endogenous MAVS andFACL4 Huh7 cells preloaded with Mitotracker. In A and C, arrows indicate MAVS on MAM or between peroxisomes and mitochondria. Single-cell images arerepresentative of at least 10 cells analyzed. Histograms display measured fluorescence intensity along the line in the merge panels, with gray lines indicatinglocal peaks of MAVS intensity. (Scale bar: 10 μm.)

Horner et al. PNAS | August 30, 2011 | vol. 108 | no. 35 | 14591

IMMUNOLO

GY

Dow

nloa

ded

by g

uest

on

Apr

il 26

, 202

1

were moderately increased within Huh7-K2040 cells, reflectingthe cellular compensation due to MAVS cleavage and RIG-Isignaling suppression (9, 10). We also conducted confocal mi-croscopy analysis and confirmed that a portion of RIG-I wasredistributed into the MAM near sites of MAVS localizationduring SenV infection (Fig. S5). Furthermore, immunoprecipi-tation analyses of the MAM fraction from PH5CH8 hepatocytesrevealed that the portion of RIG-I that redistributed to theMAM during acute RNA virus infection interacted with bothMAVS and TRAF3 (Fig. 3C). This indicates that RIG-I and itssignaling cofactor TRAF3 are distributed to the MAM duringRNA virus infection for interaction with MAVS.Contacts between the MAM and mitochondria are main-

tained by the mitochondrial fusion protein MFN2 (20). Toassess the role of MAM contacts in coordinating MAVS local-ization and signaling during virus infection, we used siRNA

technology to knock down MFN2 expression in Huh7 cells.Although MFN2 is the MAM tethering protein, it shares ho-mology with the related MFN1 protein, and both are involved inmitochondrial fusion and bind to MAVS (21, 24). Therefore, wealso used siRNA to knock down MFN1 to control for the sharedfunctions of these proteins and to specifically isolate the MAM–

mitochondrial tethering role of MFN2. Manders’ colocaliza-tion coefficient analysis (25) of confocal micrographs confirmedthat loss of MFN2 expression, but not of MFN1 expression,disrupted the ER–mitochondria interaction (Fig. S6) (20).Colocalization analysis of MAVS and TOM20 in confocalmicrographs through an object-based overlap approach (26)revealed that knockdown of MFN2, but not of MFN1, resultedin a slight but significant increase in nonmitochondrial MAVSduring SenV infection (Fig. 4 A and B and Fig. S7B). Knock-down data are provided in Fig. 5A. Colocalization analysis with

NS4A

NS3 NS3 Mito.NS3MAMNS3 Mito.

NS4A Mito.NS4AMito.

E

calnexinCox-1mEGFPPSS-1

NS3NS4A

F1 Mito

chon

dria

Mic

roso

mes

Cyt

osol

Total cell lysate

A

MA

M

Vect

or

NS

3/4A

B

HeLa cells fractionation

MAMNS4A

D

CZoom

egreM07PMPA4/3SN

NS3/4APMP70

NS3/4AMito.Mito.

HCV-K2040 replicon cells

64

64

6

37116

Merge

Merge

NS4A

NS3 Mito.

Mito.

f

Zoom

Zoom

Merge

Merge

HCV infected cells

Merge NS3/4A

%In

tens

ity%

Inte

nsity

Distance

Merge NS4A

Zoom

Merge NS3

Zoom

%In

tens

ity%

Inte

nsity

1

2

1

2

1

2

1

2

MAM Mito. MergeNS3

MAM Mito. MergeNS4A

Fig. 2. NS3/4A localizes to junctions between MAM and mitochondria. (A) Immunoblot analysis of subcellular fractions from HeLa cells stably expressingmEGFP-PSS-1 (MAM) and transiently expressing NS3/4A. (B–E) Confocal micrographs of cells transfected with mEGFP-PSS-1 or left untransfected, preloadedwith Mitotracker, and immunostained. (B) Huh7-HCV K2040 cells. White arrows indicate NS3 and NS4A adjacent to mitochondria. Histograms display themeasured fluorescence intensity along the line in the merge panels, with gray lines marking local peaks of HCV protein intensity. (C) Huh7 cells transientlyexpressing HCV NS3/4A. (D and E) Huh7.5 cells infected with HCV for 48 h. Large arrows indicate NS3 and NS4A near mitochondria, and small arrows indicatemembranous areas devoid of mitochondria. Single-cell images are representative of at least 10 cells analyzed. (Scale bars: 10 μm in B–D; 2 μm in E.)

14592 | www.pnas.org/cgi/doi/10.1073/pnas.1110133108 Horner et al.

Dow

nloa

ded

by g

uest

on

Apr

il 26

, 202

1

the peroxisomal marker PMP70 established that this increase innonmitochondrial MAVS corresponded with its enhanced as-sociation with peroxisomes (Fig. 4 C and D). In addition, itappeared that detection of peroxisomes by immunofluorescence

was easier because of an increased number or a reduced asso-ciation with mitochondria when MFN2 was knocked down (Fig.4E), perhaps explaining the increased MAVS association withperoxisomes on MFN2 knockdown.

A

CTR

L si

RN

AM

FN2

siR

NA

Merge

Mock MAVS TOM20 Merge

Zoom

Merge

C

ED

CTRL MFN1 MFN2siRNA

Pea

rson

’s C

oeffi

cien

t

0.0

0.1

0.20.3

0.4 MockSenV

Peroxisome and Mitochondria

** ***0.5

B

MAVS PMP70 TOM20 MergeZo

omMockMerge

Merge

SenV, 18h

Merge

Merge

Zoom

MAVS and Peroxisome

0.0

0.2

0.4

0.6

Loca

lizat

ion

Inde

x

MockSenV

**

CTRL MFN1 MFN2siRNA

0.00.10.20.30.40.50.60.7

Loca

lizat

ion

Inde

x

MAVS and Mitochondria

MockSenV ***

*

CTRL MFN1 MFN2siRNA

SenV, 18h

CTR

L si

RN

AM

FN2

siR

NA Zo

om

Merge

Merge

MAVS TOM20 Merge

CTR

L si

RN

AM

FN2

siR

NA

CTR

L si

RN

AM

FN2

siR

NA

MAVS PMP70 TOM20 Merge

Fig. 4. Mitochondrial–ER contacts regulate MAVS localization and signaling. (A) Immunostain of endogenous MAVS and TOM20 (OMM). Arrows indicateMAVS localized independent of TOM20. (C) Immunostain of cells expressing MAVS-mEGFP for PMP70 (peroxisomes) and TOM20. (B, D, and E) Values (mean ±SD; n = 3–4 cells) of MAVS localization to mitochondria (B) or peroxisomes (D) (object-based overlap approach) (26), or overlap between peroxisomes andmitochondria (E) (Pearson’s coefficient) (26). *P ≤ 0.05; **P ≤0.005; ***P ≤0.0005, by unpaired Student t test. Single-cell images are representative of twoexperiments and >10 cells analyzed per experiment. (Scale bar: 10 μm.)

A B

C

HCVK2040

RIG-I

NS3

MAVS

− +MAM Mito.

FL

Cytosol− + − +− +

Total

65

116

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8 9 10NS3

64

1168237

82

37

26

65

tubulin

calnexinCox-1

FACL4

Pex19

MA

MM

itoch

ondr

iaM

icro

som

esC

ytos

ol

Huh7

MA

MM

itoch

ondr

iaM

icro

som

esC

ytos

ol

MA

M &

Pex

HCV-K2040

116 RIG-I

MA

M &

Pex

IP: RIG-I

64

82

116

*

MAM CytosolSenV

RIG-I

MAVS

TRAF3

*

IB− + − +

InputFACL4

tubulin64

82

C

Fig. 3. The MAM is an innate immune signaling platform. (A and B) Immunoblot analysis of fractionated Huh7 (−) and Huh7-HCV K2040 (+) cells. Arrowsindicate full-length (FL) and cleaved (C) MAVS. (C) Immunoblot analysis of fractions immunoprecipitated with anti–RIG-I from mock- or SenV-infected (for 8 h)PH5CH8 cells and input. *Nonspecific band. Fractionation markers: calnexin, ER; Cox-1, mitochondria; FACL4, MAM; tubulin, cytosol; Pex19, peroxisomes.

Horner et al. PNAS | August 30, 2011 | vol. 108 | no. 35 | 14593

IMMUNOLO

GY

Dow

nloa

ded

by g

uest

on

Apr

il 26

, 202

1

In addition to increasing MAVS localization with perox-isomes, treatment of cells with siRNA specific to MFN2 or withmethyl-β-cyclodextrin, which disrupts MAM–mitochondria con-tacts through distinct mechanisms (27), enhanced RIG-I pathwaysignaling as defined by an increase in phosphorylated active IRF-3, enhanced IFN-β promoter activity, and ISG56 expression afterSenV infection (Fig. 5 A and B). We note that knockdown ofMFN2 enhanced RIG-I signaling despite an attenuation of thepreviously described virus-induced mitochondrial elongation(Fig. S7A), demonstrating that its role in the antiviral responseis actually independent of a role in mitochondrial elongation(28). MFN1 knockdown attenuated SenV-induced RIG-I sig-naling, as described previously (24, 28), while maintaining ER–mitochondria contacts (Fig. S6). When infected with HCV,MFN2 knockdown of cells resulted in a moderate but significantreduction in HCV RNA early after infection up to the point ofMAVS cleavage by NS3/4A, consistent with acute enhancementof innate immune defenses induced by RIG-I–HCV RNA in-teraction due to deregulated RIG-I pathway signaling (Fig. 5C)(7, 29). Indeed, MFN2 knockdown reduced the permissivenessof cells for infection by HCV and vesicular stomatitis virus(VSV), consistent with enhanced innate immune defenses (Fig.5D). Therefore, MAVS localization and antiviral signaling aresupported by the MAM–mitochondrial tethering function ofMFN2, which may regulate MAVS distribution among mito-chondria, MAM, and peroxisomes.We also found that during SenV infection, the interaction

between peroxisomes and mitochondria increased in an MFN2-dependent manner (Fig. 4 C and E). This observation suggeststhat MAM–mitochondrial tethering mediates an increased in-teraction between these organelles during RIG-I pathway acti-vation. Indeed, confocal microscopy of SenV-infected Huh7 cellsrevealed that the MAM forms a synapse between peroxisomesand mitochondria (Fig. 5E). An apparent synapse between per-oxisomes and mitochondria anchored by MAVS also can bedetected on MAVS overexpression (Fig. 1A). Moreover, we

found that this synapse is targeted by the NS4A subunit of theHCV protease complex (Fig. 5F). These observations raise thepossibility that NS3/4A targeting of this MAM-anchored synapsemay regulate innate immune signaling by MAVS from MAM,peroxisomes, and mitochondria.

DiscussionOur results identify an innate immune signaling synapse in whichthe MAM serves as the central scaffold that coordinates MAVS-dependent signaling of the RIG-I pathway between mitochondriaand peroxisomes. This model implicates the MAM in immunesignaling to a variety of viruses in addition to HCV that are en-gaged by RIG-I–like receptors to trigger the immune response(2). TheMAM-anchored synapse contains both MAVS and RIG-I pathway factors that physically interact in RNA virus-infectedcells to drive the immune actions of an MAVS signalosome.STING, a putative RIG-I signaling cofactor and IPS-1 bindingprotein, also may have MAM localization, but this localization islikely specific to cell type and context (30). Because the viral NS3/4A protease targets the MAM-anchored synapse and cleavesMAVS from the MAM, but not from mitochondria, to ablateinnate immune signaling during HCV infection, the MAM-resi-dent MAVS likely mediates the RIG-I pathway function againstHCV. This notion is further supported by previous studiesshowing that cells supporting chronic HCV RNA replication(which, as we show here, contain intact mitochondrial-localizedMAVS but not intact MAM-localized MAVS) are unable tosignal IRF-3 activation and the expression of IRF-3 target genesdespite the presence ofMAVS on themitochondria (9). AlthoughMAM-localized MAVS directs innate immunity against HCV,our observations do not exclude a role for mitochondrial MAVSin innate immune signaling. In this respect, mitochondrial MAVSsupports apoptotic signaling of innate immune actions againstvirus infection (31, 32), and also could serve as a local source ofMAVS for distribution among peroxisomes and MAM throughdynamic membrane trafficking regulated by MFN2. This latter

A

MFN1

Tubulin

IRF-3

P-IRF-3SenV

ISG56

MFN2

CTRL MFN1 MFN2siRNA

- +SenV - + - +MBCD- +

DC

E

F

B

0

10

20

30

40IFN- promoter activity

CTRL MFN1 MFN2 MBCDsiRNA

Rel

ativ

e lu

cife

rase

uni

ts

***

150 M

*

**

MockSenV

HC

V+

cells

/2x1

0^3

plat

ed

CTRLMFN1 MFN2siRNA

020406080

100120 **

HCV infection

0

10

20

30

% V

SV

+ ce

lls

CTRL MFN1 MFN2

siRNA

******

VSV infection

0

1

2

3

4

5

6

0 hpi 24 hpi 48 hpi 72 hpi

10^6

HC

V c

opie

s /

25

g to

tal R

NA

*

siRNA

CTRLMFN1MFN2

MAVS

HCV NS5Atubulin

FLC

0 24 48 72 hpi

Merge NS4A PMP70 TOM20 Merge

Zoom

SenV, 18h

Merge PSS-1 PMP70 TOM20 Merge

Zoom

SenV, 18h 3D reconstruction

PMP70 TOM20 PSS-1

PMP70 TOM20 NS4A

μ

β

Fig. 5. Mitochondrial–ER contacts regulate RIG-I pathway antiviral signaling. (A and B) Immunoblot analysis and IFN-β promoter luciferase expression ofHuh7 cells treated with nontargeting control (CTRL), MFN1, or MFN2 siRNA pools or methyl-β cyclodextrin and then mock- or SenV-infected. Values aremean ± SD (n = 3) of one of three representative experiments. *P ≤ 0.05; **P ≤0.005; ***P ≤0.0005, by unpaired Student t test. (C) HCV copy number (mean ±SD; n = 3). (Inset) Immunoblot analysis of MAVS; full-length (FL) and cleaved (C). (D) (Upper) HCV-positive cells after 72 h of infection (mean ± SD; n = 5–6).(Lower) Percent VSV-positive cells after 16 h of VSV infection (mean ± SD; n = 3–5, >1000+ cells counted). (E and F) Confocal micrographs of immunostainedSenV-infected Huh7 cells expressing mEGFP-PSS-1 (MAM) (E) or HCV NS4A-mEGFP (F) with PMP70 (peroxisomes) and TOM20 (OMM) and 3D reconstructionof Z stacks. Arrows mark synapses. Single-cell images are representative of two experiments and >10 cells analyzed per experiment. (Scale bar: 10 μm.)

14594 | www.pnas.org/cgi/doi/10.1073/pnas.1110133108 Horner et al.

Dow

nloa

ded

by g

uest

on

Apr

il 26

, 202

1

idea is supported by previous work showing that peroxisomes arenot required for populating mitochondria with MAVS (6), sug-gesting that peroxisomes may receive MAVS by membrane traf-ficking through MAM–mitochondrial interactions.We found that when MAM–mitochondrial contacts were dis-

rupted, there was increased localization of endogenous nativeMAVS to peroxisomes, with a concomitant increase in signalingto IFN-β compared with cells maintaining MAM–mitochondriacontacts and harboring less peroxisomal MAVS. In contrast,previous work has shown that ectopic MAVS targeted solely toperoxisomes through replacement of the native MAVS trans-membrane domain with a peroxisome-targeting domain does notsignal to IFN-β (6). In that case, the presence of the nonnativeMAVS transmembrane domain likely precluded essential mem-brane interactions with signaling factors that drive IFN-β pro-duction, thus explaining the divergent findings. In the presentstudy, peroxisomal MAVS enhanced signaling to IFN-β, possiblybecause when localized to peroxisomes, MAVS no longer inter-acts with mitochondrial negative regulatory proteins of innateimmune signaling, including NLRX1 (33). In support of this,NLRX1 has been shown to be unable to suppress signaling byperoxisomal MAVS (6). Placement of NLRX1 or of other neg-ative regulators of RIG-I signaling on mitochondria also mightexplain why MAVS must be redistributed to MAM to initiateinnate immune signaling. In this case, the mitochondria-to-MAM-to-peroxisome redistribution of MAVS would release it fromsignaling inhibition by a negative regulator, and the MAM wouldprovide the platform for recruiting positive signaling effectorproteins to interact with MAVS for innate immune signaling.Our study and others indicate that dynamic immune signaling is

coordinated by theMAMthroughmembrane–protein interactionsgoverned by MFN2, with membrane tethering sites acting as sig-naling microdomains that contain localized sources of MAVS andother signaling factors (15). This organization is analogous to the

immunologic synapse that forms between a T cell and an antigen-presenting cell in which adhesion molecules tether opposingplasma membranes to establish T-cell receptor signaling micro-domains supplied with key signaling factors through membraneinteractions (34). Taken together with the role of the MAM inNLRP3 inflammasome signaling (16), our findings indicate thattheMAMplays a central role in initiating both the innate immuneand inflammatory responses to infection. Therapeutic strategiesdirected toward regulating MAM processes could serve to pre-serve and enhance antiviral immunity.

Materials and MethodsAdditional information is provided in SI Materials and Methods.

Cell Culture and Viruses. Huh7, Huh7.5, and Huh7-HCV K2040 replicon cellsthat propagate culture-adapted variants of the Con1 HCV (genotype 1B)subgenomic replicon RNA (3), along with immortalized human hepatocytePH5CH8 cells, were cultured according to standard techniques. SenV strainCantell was obtained from Charles River Laboratory. Cell culture adaptedHCV JFH1 genotype 2A strain was propagated and infectivity titrated, asdescribed previously (35). VSV-GFP infections were done at MOI 1 for 16h.Methyl-β-cyclodextrin treatment used a concentration of 150 μm in DMSO.

Subcellular Fractionation.MAM, mitochondria, and microsomes were isolatedfrom cells using Percoll gradient fractionation (18). Equivalent amounts ofprotein from each fraction were analyzed by SDS/PAGE and immunoblot.

ACKNOWLEDGMENTS. We thank our colleagues for reagents, includingA. Colberg-Poley for pmEGFP-huPSS-1; the W. M. Keck Center for AdvancedStudies in Neural Signaling for microscopy assistance; K. Chang for technicalassistance; and D. Stetson, M. Diamond, and members of the M.G. laboratoryfor discussion and a critical reading of the manuscript. This work wassupported by the National Institutes of Health and the Burroughs WellcomeFund. S.M.H. is a fellow of the Irvington Institute Fellowship Program ofthe Cancer Research Institute.

1. Kumar H, et al. (2006) Essential role of IPS-1 in innate immune responses against RNAviruses. J Exp Med 203:1795–1803.

2. Loo YM, et al. (2008) Distinct RIG-I and MDA5 signaling by RNA viruses in innateimmunity. J Virol 82:335–345.

3. Sumpter R, Jr., et al. (2005) Regulating intracellular antiviral defense and permis-siveness to hepatitis C virus RNA replication through a cellular RNA helicase, RIG-I.J Virol 79:2689–2699.

4. Yoneyama M, et al. (2004) The RNA helicase RIG-I has an essential function in double-stranded RNA-induced innate antiviral responses. Nat Immunol 5:730–737.

5. Seth RB, Sun L, Ea CK, Chen ZJ (2005) Identification and characterization of MAVS,a mitochondrial antiviral signaling protein that activates NF-κB and IRF 3. Cell 122:669–682.

6. Dixit E, et al. (2010) Peroxisomes are signaling platforms for antiviral innate immunity.Cell 141:668–681.

7. Saito T, Owen DM, Jiang F, Marcotrigiano J, Gale M, Jr. (2008) Innate immunity in-duced by composition-dependent RIG-I recognition of hepatitis C virus RNA. Nature454:523–527.

8. Foy E, et al. (2003) Regulation of interferon regulatory factor-3 by the hepatitisC virus serine protease. Science 300:1145–1148.

9. Foy E, et al. (2005) Control of antiviral defenses through hepatitis C virus disruptionof retinoic acid–inducible gene-I signaling. Proc Natl Acad Sci USA 102:2986–2991.

10. Loo YM, et al. (2006) Viral and therapeutic control of IFN-β promoter stimulator 1during hepatitis C virus infection. Proc Natl Acad Sci USA 103:6001–6006.

11. Meylan E, et al. (2005) Cardif is an adaptor protein in the RIG-I antiviral pathway andis targeted by hepatitis C virus. Nature 437:1167–1172.

12. Moradpour D, Penin F, Rice CM (2007) Replication of hepatitis C virus. Nat RevMicrobiol 5:453–463.

13. Vance JE (1990) Phospholipid synthesis in a membrane fraction associated withmitochondria. J Biol Chem 265:7248–7256.

14. Lebiedzinska M, Szabadkai G, Jones AW, Duszynski J, Wieckowski MR (2009) Inter-actions between the endoplasmic reticulum, mitochondria, plasma membrane andother subcellular organelles. Int J Biochem Cell Biol 41:1805–1816.

15. Hayashi T, Rizzuto R, Hajnoczky G, Su TP (2009) MAM: More than just a housekeeper.Trends Cell Biol 19:81–88.

16. Zhou R, Yazdi AS, Menu P, Tschopp J (2011) A role for mitochondria in NLRP3 in-flammasome activation. Nature 469:221–225.

17. Hoepfner D, Schildknegt D, Braakman I, Philippsen P, Tabak HF (2005) Contribution ofthe endoplasmic reticulum to peroxisome formation. Cell 122:85–95.

18. Bozidis P, Williamson CD, Colberg-Poley AM (2007) Isolation of endoplasmic re-ticulum, mitochondria, and mitochondria-associated membrane fractions from

transfected cells and from human cytomegalovirus-infected primary fibroblasts. CurrProtoc Cell Biol chap 3:unit 3.27.

19. Stone SJ, Vance JE (2000) Phosphatidylserine synthase-1 and -2 are localized to mi-tochondria-associated membranes. J Biol Chem 275:34534–34540.

20. de Brito OM, Scorrano L (2008) Mitofusin 2 tethers endoplasmic reticulum to mito-chondria. Nature 456:605–610.

21. Yasukawa K, et al. (2009) Mitofusin 2 inhibits mitochondrial antiviral signaling. SciSignal 2:ra47.

22. Brass V, et al. (2008) Structural determinants for membrane association and dynamicorganization of the hepatitis C virus NS3-4A complex. Proc Natl Acad Sci USA 105:14545–14550.

23. Lin R, et al. (2006) Dissociation of a MAVS/IPS-1/VISA/Cardif-IKKε molecular complexfrom the mitochondrial outer membrane by hepatitis C virus NS3-4A proteolyticcleavage. J Virol 80:6072–6083.

24. Onoguchi K, et al. (2010) Virus-infection or 5′ppp-RNA activates antiviral signalthrough redistribution of IPS-1 mediated by MFN1. PLoS Pathog 6:e1001012.

25. Manders EM, et al. (2003) Four-dimensional imaging of chromatin dynamics duringthe assembly of the interphase nucleus. Chromosome Res 11:537–547.

26. Bolte S, Cordelières FP (2006) A guided tour into subcellular colocalization analysis inlight microscopy. J Microsc 224:213–232.

27. Sano R, et al. (2009) GM1-ganglioside accumulation at the mitochondria-associatedER membranes links ER stress to Ca(2+)-dependent mitochondrial apoptosis. Mol Cell36:500–511.

28. Castanier C, Garcin D, Vazquez A, Arnoult D (2010) Mitochondrial dynamics regulatethe RIG-I–like receptor antiviral pathway. EMBO Rep 11:133–138.

29. Uzri D, Gehrke L (2009) Nucleotide sequences and modifications that determineRIG-I/RNA binding and signaling activities. J Virol 83:4174–4184.

30. Ishikawa H, Ma Z, Barber GN (2009) STING regulates intracellular DNA-mediated, typeI interferon-dependent innate immunity. Nature 461:788–792.

31. Holm GH, et al. (2007) Retinoic acid–inducible gene-I and interferon-β promoterstimulator-1 augment proapoptotic responses following mammalian reovirus in-fection via interferon regulatory factor-3. J Biol Chem 282:21953–21961.

32. Yu CY, Chiang RL, Chang TH, Liao CL, Lin YL (2010) The interferon stimulator MAVSfacilitates cell death by disrupting the mitochondrial membrane potential and byactivating caspases. J Virol 84:2421–2431.

33. Moore CB, et al. (2008) NLRX1 is a regulator of mitochondrial antiviral immunity.Nature 451:573–577.

34. Yokosuka T, Saito T (2010) The immunological synapse, TCR microclusters, and T cellactivation. Curr Top Microbiol Immunol 340:81–107.

35. Kato T, et al. (2006) Cell culture and infection system for hepatitis C virus. Nat Protoc1:2334–2339.

Horner et al. PNAS | August 30, 2011 | vol. 108 | no. 35 | 14595

IMMUNOLO

GY

Dow

nloa

ded

by g

uest

on

Apr

il 26

, 202

1